Talks
Ostensibly done
Talks involving the same speaker or co-presenter can not be scheduled simultaneously, and preferably should be well-spaced.
Talks on the same topic should not be scheduled simultaneously. Preferably they should be sequential, in the same room.
Do not schedule 'withdrawn' talks. In fact, ONLY schedule 'accepted' talks. (check that there are no talks in any other state)
'Improving client-side HTTP' aims to elicit discussion, so should not be scheduled against something heavyweight, he needs relevant people to be present in order to succeed. LAC: "the way I would implement this is to schedule his talk FIRST and give him the very first pick of the others to schedule against. because I think that this is the very most important talk to schedule as best as we can"
Both talks by Tobias Rundström (Corporate FOSS and OSS, BSS and Python) can only be scheduled on wed 1st or thu 2nd.
'Twisted, AMQP and Thrift' is a component of FluidDB, don't schedule against 'Introducing FluidDB'
Terry Jones ('Introducing FluidDB') would like to fly out Wed night, and so be scheduled Tue or Wed.
Tutorial Pyjamas: Simple RIA Web apps is to be scheduled for the Monday.
'PyPy - complete and fast', is 60 mins, could alternatively be 2x 30 mins
Talks of different lengths should not be scheduled simultaneously (eg. one of 90 mins against 2 of 45 mins) if it is at all possible to avoid. The disruption is not worth it, and delegates leaving one talk can't enter the other halfway through, so that time is wasted for them.
Talks which introduce a sprint should not be scheduled simultaneously. The wiki indicates the only talks to which this applies are: PyPy, Turbogears(Ramm), Pyjamas and Grok.
The hotpy talks should follow one after the other. They really are one long talk with a pause in the middle.
Zen of Web should be scheduled early, in case there is somebody new to web programming who should see this talk first.
Luke Leighton's stuff probably needs to go together, because if you want to hear it you probably want to hear all of them, and if you do not, you want to hear neither.
Russel Winder's GIL isn't Evil should be scheduled 'on the flat', preferably in the Recital Hall. (alternatively the ABH, or Arena Foyer.)
I thought we agreed Mike Dirolfs MongoDB talk should be 45 minutes.
Zeth's talk is about my XML module for Django, it is really a django talk. I am not sure that education has much to do with it
The iron python talks are all spread out. is it possible to group them together?
I am not sure, but I think that Python and Nuke and the Visual studio plugin should not be scheduled against each other.
Not Done
Luke Leighton's two 'proofread' talks will become one. Schedule them next to Jonathan Fine's javascript. they are one talk but not scheduled next to JF. why was that a considered a good idea? Laura
Russel Winder's GIL isn't Evil should be scheduled 'on the flat', preferably in the Recital Hall. (alternatively the ABH, or Arena Foyer.) (Why is this? Laura)
Samuele's Javascript would naturally be scheduled in line with Luke's Pyjamas stuff -- I think this is wrong. Laura
Ctypes should not be scheduled vs christian tismer's talk on psyco. but they should be scheduled with ironclad, in the same session, ideally ironclad followed by ctypes. They are aimed at exactly the same audience. (half done, the whole was impossible to implement and also make a nice Iron Python track.)
jumping worlds with python should not be scheduled against python .net and the weather. They should follow each other in one session. they are about the same thing, and indeed the authors of both need to meet each other. (Again, conflicts with the iron python track idea. the talks were not suitable in length either, if they had been 45 minutes long it would have been easier in some ways to schedule them)